Ollako vai eikö olla - eksistentiaalilauseen subjektin kohtalonkysymys

• Marja-Liisa Helasvuo

Avainsanat: puheen syntaksi, puhuttu kieli (ks. myös puhekieli), referenssi, syntaksi

Abstrakti

To be or not to be - the question of existential subject in Finnish (englanti)

3/1996 (100)

Marja-Liisa Helasvuo

To be or not to be - the question of existential subject in Finnish

There is a wealth of research on so-called existential clauses (henceforth: e-clauses) in Finnish. This article examines the nominal constituents in e-clauses from the viewpoint of their use in conversational discourse. Under discussion are contructions which instantiate the following schema:

NP[loc] + V[exist.] + NP[nom/ptv] (e-NP)

Schemas are assumed to be conventionalized patterns of language use, of which e-clauses represent one example. The aim of the article is to investigate why this particular pattern is needed in discourse. I have approached this question by looking at what kind of referents e-clauses bring into discourse and how these referents are treated after being introduced. In the literature, the syntactic function of the postverbal NP (henceforth: e-NP) has been under dispute, but according to the prevalent view, it is a subject. This paper compares the e-NP to both the subject and the object and shows that the discourse functions it serves are quite different from those of subject and object. It also differs from subject and object in its coding properties.

For this study, I collected data from six conversations between speakers of Finnish. From this data I extracted smaller segments for closer analysis, together comprising a database of approx. 40 minutes of spoken interaction. In the database, there were 85 e-clauses (constructions with a locative NP + V + e-NP).

It has been proposed that e-clauses function to introduce new referents into discourse in the e-NP role (Whmki 1975). This claim is also supported by the present study: the majority of e-NPs refer to referents new to the discourse and these referents are mainly non-human. These information flow features (i.e. preference for new mentions and non-human referents) characterize not only e-NPs but also NPs in the object role. In contrast, NPs in the subject role refer mainly to referents that have already been mentioned in the discourse. Subject NPs usually refer to human referents which are being tracked in the discourse. Thus, subject NPs lack certain information flow features that object NPs and e-NPs share.

However, if we study how the referents are treated in the discourse after being introduced, we can see an interesting pattern: referents that are introduced into the discourse in the e-NP role are usually not tracked in the discourse; instead, they are dropped after one mention in the e-NP role. In other words, they do not seem to be central participants in the discourse. In contrast, referents that are first introduced in the object role are usually mentioned several times during the discourse. If referents first mentioned in the e-NP role are usually not mentioned again, why are they introduced into the discourse at all?

Huumo (1996) looks at e-clauses from the viewpoint of the semantics of the locative NP. According to him, an e-construction "states a predication about the locative space by reporting its content". From a discourse point of view we could say that the locative NP is the starting point (Chafe 1994) or reference point (Langacker 1995) of the construction from which the predication proceeds. The locative NP usually refers to a referent which has already been mentioned in the discourse. In an e-clause, a relationship is established between the locative space and a new referent expressed in the e-NP role. This paper argues that e-NPs serve to characterize the referent of the locative NP by situating a new referent in the locative space referred to by the locative NP. Thus, e-NPs serve to characterize locative space rather than introduce new participants into the discourse. It is only natural, then, that the referents of e-NPs are usually not tracked in the discourse, but mentioned only once. In this respect e-NPs differ from NPs in the core roles, from both subjects and objects.

In Finnish linguistics, e-NPs have traditionally been analyzed as subjects. There are problems with this view and these have also been discussed in the literature. Despite the problems, there has been a preference for rather a broad definition of subject that includes inter alia e-NPs. This article argues for the alternative view which assumes that only nominative NPs which can trigger agreement with the verb are subjects. This narrow definition of subject allows us to give a more unified account of the syntactic and information flow characteristics of NPs in the subject role. As these characteristics are not shared by e-NPs, e-NPs should not be regarded as subjects.

To summarize, this article shows that although e-clauses do introduce new referents into the discourse, the referents are not central to the discourse, but instead, they are mentioned only once. What seems more central is the locative NP: it is the one that is being tracked in the discourse. It serves as the starting point of the clause. In an e-clause, the locative space referred to by the locative NP is characterized by relating to it a new referent that is expressed by the e-NP.

There has been a long-standing dispute over the syntactic function of the e-NP. The subject role as defined in this paper is characterized by syntactic features that are independent of clause type. In contrast, in order to understand the grammaticization of the e-NP, reference has to be made to the notion of clause type: the whole clause type has been grammaticized. In an e-clause, the case marking and position of the e-NP have been crystallized as well as the choice of the verb (olla 'to be' and certain other verbs) and the position and case marking of the locative NP.